“[A]n indispensable factor to liability founded upon negligence is the existence of a duty of care owed by the wrongdoer to the person injured . . . .” {Routh v. Quinn (1942) 20 Cal.2d 488, 491 [127 P.2d 1, 149 A.L.R. 215].) I concur with my colleagues that there is no duty in this case chiefly because I do not believe University of California (the University) placed plaintiff in a position of danger or materially contributed to the harm she suffered; nor do I believe the University should reasonably have anticipated plaintiff would rely on it to protect her against the harm she suffered.
In seeking to establish a duty, plaintiff emphasizes that the University either knew or should have known of the high risk of sexual assaults upon its female students. In her brief she sums up her contention as follows: “[T]he information available to respondents at or about the time plaintiff was raped, in mass media, social science research, and professional educators’ publications, reveals that by the early to mid-1980’s, the danger to young women in plaintiff’s position of precisely the injury she alleges—gang rape by male student acquaintances—had been clearly documented in publications of which respondents could not reasonably have been unaware.”1
*443It is indisputable that a large body of recent empirical studies documents an alarming level of sexual assaults against female students at American colleges and universities. In one of the most impressive studies, based on a national sample of 6,159 male and female students enrolled in 32 institutions of higher education across the United States, 12.1 percent of the women respondents said they had been the victim of at least 1 attempted rape; 15.4 percent said they actually had been raped; and 53.7 percent said they had experienced some form of “sexual victimization.” (Koss et al., The Scope of Rape: Incidence and Prevalence of Sexual Aggression and Victimization in a National Sample of Higher Education Students, 55 J. of Consulting & Clinical Psych. 162, 166 (1987).) Other studies report similar findings. (See, e.g., Rapaport & Burkhart, Personality and Attitudinal Characteristics of Sexually Coercive College Males, 93 J. of Abnormal Psych. 216, 220 (1984); Lott et al., Sexual Assault and Harassment: A Campus Community Case Study, 8 Signs: J. of Women in Culture & Society 296, 306 (1982); Fenstermacher, Acquaintance Rape on Campus: Responsibility and Attributions of Crime, in Pirog-Good & Stets (edits.) Violence In Dating Relationships: Emerging Social Issues, at pp. 257-271 (1989).)
However, none of the studies just cited, nor others brought to our attention by plaintiff, suggest that the incidence of sexual assault occurring on college and university campuses is any different from that occurring elsewhere in American society or that female college or university students are a particularly vulnerable group because they reside on or near a campus or due to the consumption of alcohol in campus dormitories. The reason many studies of the prevalence of rape and other forms of sexual aggression have involved college students is simply because they are an available sample “in the same age range as the bulk of [all] rape victims and offenders. The victimization rate for women peaks in the 16-19 year-old age group, and the second highest rate occurs in the 20-24 year-old age group. The victimization rates for these [age] groups are approximately 4 times higher than the mean for all women. Also, 45% of all alleged rapists who are arrested are individuals under age of 25.” (Koss et al., The Scope of Rape, supra, 55 J. of Consulting & Clinical Psych, at p. 163, citations omitted.)
Appallingly, sexual aggression is endemic throughout our society. Finding that “the locus of violence against women rests squarely in the middle of what our culture defines as ‘normal’ interaction between men and women,” one study asserts that “the average American woman is just as likely to suffer a sexual attack as she is to be diagnosed as having cancer, or to experience a divorce.” (Johnson, On the Prevalence of Rape in the United States, 6 Signs: J. of Women in Culture & Society 136, 146 (1980), see also, Brownmiller, Against Our Will: Men, Women and Rape (1975); Katz & *444Mazur, Understanding the Rape Victim: A Synthesis of Research Findings (1979).)
This information is of no assistance to plaintiff. While the studies she relies upon show that women her age are particularly vulnerable to sexual assault, they establish no connection between sexual violence and the conduct of university administrators (or other landlords); indeed, the empirical data strongly suggests there is no such connection.
Nor do the studies or any other evidence provide reason to think university students are unaware of the risk of sexual violence on campus and elsewhere or that they rely for protection on university administrators. To suppose that it is uniquely within the power of campus officials to diminish this risk would not only exaggerate their abilities but trivialize the problem.
Though it involved a cause of action for a dangerous condition of property under Government Code section 835, and a different type of premises, Hayes v. State of California (1974) 11 Cal.3d 469 [113 Cal.Rptr. 599, 521 P.2d 855] is nonetheless relevant. In that case, two youths were attacked and beaten by unknown third persons while asleep at night on a beach on the campus of the University of California at Santa Barbara. A unanimous Supreme Court found the Regents of the University and other government defendants were under no duty to warn against criminal conduct. The court acknowledged “that the warning called for by plaintiffs might be beneficial in some instances . . . (Id., at p. 472, fn. omitted.) But because of “public awareness of the prevalence of crime and policy factors,” the court believed “it would serve little purpose for government to further remind the public of this unfortunate circumstance in society.” (Id., at p. 473.) The court also felt that warning of the danger of criminal conduct on university premises—“unlike cautioning against a specific hazard in the use of property—admonishes against any use of the property whatever, thus effectively closing the area.” (Ibid., italics in original.)
As explained in the majority opinion, plaintiff’s premises liability theory is untenable. Therefore, like Hayes, the present case is distinguishable from cases, such as Peterson v. San Francisco Community College Dist. (1984) 36 Cal.3d 799 [205 Cal.Rptr. 842, 685 P.2d 1193], where the duty to warn was imposed in connection with “physical defects” of property that increased the risk of crime. (Id., at p. 813; see also Kenny v. Southeastern Pennsylvania Transp. (3d Cir. 1978) 581 F.2d 351, cert. den. 439 U.S. 1073 [59 L.Ed.2d 39, 99 S.Ct. 845].) It is also different from cases in which the landlord had notice of repeated criminal assaults in the particular building *445in which the plaintiff resided and of the likelihood of repeated attacks, and took no precautions to protect his tenants or concealed the problem. (See, e.g., O'Hara v. Western Seven Trees Corp. (1977) 75 Cal.App.3d 798, 802 [142 Cal.Rptr. 487]; Kline v. 1500 Massachusetts Avenue Apartment Corp. (D.C. Cir. 1970) 439 F.2d 477, 479, 481.)
Absent a special relationship, a person who has not created a peril may not be held liable for failure to protect against it. (Williams v. State of California (1983) 34 Cal.3d 18, 23 [192 Cal.Rptr. 233, 664 P.2d 137].) Nor can the University be said to be under a legal duty to protect plaintiff from harm simply because it adopted a policy forbidding the consumption of alcohol in its dormitories and took some uneven action to enforce that policy. (See Seavey, Reliance Upon Gratuitous Promises or Other Conduct (1951) 64 Harv. L. Rev. 913, 918, 919, fn. 24 [“a gratuitous promise should not acquire validity merely by partial performance.”]
Plaintiff has not only failed to establish a close connection between the university’s conduct and the injury suffered by plaintiff, which is a factor bearing upon the existence of the asserted “duty” (Rowland v. Christian (1968) 69 Cal.2d 108, 113 [70 Cal.Rptr. 97, 443 P.2d 561, 32 A.L.R.3d 496]), but she ignores the legal significance of the fact that the conduct she complains of constitutes a nonfeasance rather than a misfeasance. Where, as here, the defendant’s conduct made the plaintiff’s situation no worse, and merely failed to benefit her by interfering in her affairs, liability is imposed only where the defendant possesses considerably more power over the plaintiff’s welfare than is evident here. (Prosser & Keeton, Torts (5th ed. 1984) pp. 373-374; Harper & Kime, The Duty to Control the Conduct of Another (1934) 43 Yale L J. 886; Bohlen, The Moral Duty to Aid Others as a Basis of Tort Liability (1908) 56 U. Pa. L. Rev. 217, 219.)
Plaintiff has also oversimplified the issue of foreseeability. The question in this case is not simply whether sexual violence on campus was foreseeable— which I think must be conceded—but whether plaintiff’s reliance on the ability of the University to protect her against sexual assault was both reasonable and foreseeable and whether the University should also have foreseen an unreasonable likelihood of harm as a result of such reliance, if the University’s promise to supervise student conduct in the dormitory was not carried out. Plaintiff does not satisfactorily address these requirements. For example, in connection with the element of reliance, she does not allege what precaution she would otherwise have taken to prevent the sexual assaults she experienced, which, as indicated, are not confined to University premises.
*446The imposition of a duty to exercise care with resulting liability for breach would hold the University liable for a risk it neither created nor exacerbated nor can readily abate. As my colleagues point out, such a duty cannot be imposed without resurrecting the university’s role of in loco parentis, which is no longer feasible, even accepting the doubtful assumption it would be wise. (Baldwin v. Zoradi (1981) 123 Cal.App.3d 275, 287-288 [176 Cal.Rptr. 809]; Bradshaw v. Rawlings (3d Cir. 1979) 612 F.2d 135, 138-140.)
For the foregoing reasons, in addition to those set forth in the majority opinion, I agree that the judgment below should be affirmed.
My colleagues are unwilling to take judicial notice of these numerous studies and articles because “it is impossible for us to determine their scientific validity or relevance.” (Maj. opn., ante, at p. 440, fn. 1.) I do not share this concern. Plaintiif claims the academic studies and other reports showing high levels of sexual violence on campus were so numerous that university administrators could not claim this danger was unforeseeable, entirely apart from the “scientific validity” of any particular study. Furthermore, defendant University has not claimed that any of the cited studies are scientifically invalid or factually misleading.